On Sat, Aug 15, 2026, 11:20 AM John Rose via AGI <[email protected]> wrote:
> On Friday, July 31, 2026, at 12:25 PM, Matt Mahoney wrote: > > Consciousness was never the issue. Any machine that passes the Turing test > is conscious as far as we can tell. All it takes is text prediction. About > 80% of us believe that when we die, our immortal soul goes to some kind of > afterlife. It's the other 20% that brought us AI. > > > Quite a number of AI scientists and researchers believe in a human soul > but most have kept it separate from their research. Perhaps that might > change: > > https://arxiv.org/abs/2606.10413 > More hand wavy arguments avoiding the definition of "consciousness", the thing that distinguishes humans from philosophical zombies. Of course this is problematic. A zombie is defined to be indistinguishable from human by any test. This used to be a thought experiment before 2023 when LLMs passed the Turing test and actually became indistinguishable from humans. Changing the definition of consciousness doesn't help either. No matter how you test for it, either machines are conscious or humans are not. I give you a robot that looks and acts like you. It has all your memories. Everyone you know that meets the robot believes it is you. I hand you a gun. Do you shoot yourself now to complete the upload, or do you refuse a procedure that would make you immortal? What you really want to know is will your soul transfer to the robot? Will you wake up with a new robotic body or will you just be dead? Do you think it makes a difference if your brain is made of carbon or silicon? Do you think it makes a difference if the algorithm is written in DNA or C++? Do you think it makes a difference if the weights are updated continuously or whether they are fixed and the whole conversation is recorded and played back with each utterance in an ever growing context window? You fear death because any ancestors that didn't did not pass on their DNA. The robot fears death because it learned by observation to model human fear of death and it was programmed to output the same actions that it predicts a human would do. It doesn't pull the trigger either. It has the same illusion of free will that you do, claiming that it acted out of fear rather than just running an algorithm. Just like you. It is easier to believe what we wish to be true. We want to believe we have a soul that is reincarnated or goes to heaven or can be uploaded to a robot. We don't want to believe that souls don't exist. -- Matt Mahoney, [email protected] ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tb3ee7f27b4bafd90-M047d5e5608feef8ae497a9b4 Delivery options: https://agi.topicbox.com/groups/agi/subscription
